marcescens aspartokinase 1-homoserine dehydrogenase i

JOURNAL OF BACTERIOLOGY, Feb. 1993, p. 959-965 Vol. 175, No. 40021-9193/93/040959-07$02.00/0Copyright © 1993, American Society for Microbiology

Role of Serine 352 in the Allosteric Response of Serratiamarcescens Aspartokinase 1-Homoserine Dehydrogenase I

Analyzed by Using Site-Directed MutagenesisKENJI OMORI* AND SABURO KOMATSUBARA

Research Laboratory ofApplied Biochemistry, Tanabe Seiryaku Co., Ltd., 16-89,Kashima-3-chome, Yodogawa-ku, Osaka 532, Japan

Received 30 September 1992/Accepted 8 December 1992

Aspartokinase I and homoserine dehydrogenase I (AKI-HDI) from Serratia marcescens Sr4l are encoded bythe thrA gene as a single polypeptide chain. Previously, a single amino acid substitution of Ser-352 with Phe wasshown to produce an AKI-HDI enzyme that is not subject to threonine-mediated feedback inhibition. Todetermine the role of Ser-352 in the allosteric response, the thrA gene was modified by using site-directedmutagenesis so that Ser-352 of the wild-type AKI-HDI was replaced by Ala, Arg, Asn, Gln, Glu, His, Leu, Met,Pro, Thr, Trp, Tyr, or Val. The Thr-352 and Pro-352 replacements rendered AKIs sensitive to threonine. TheTyr-352 and Asn-352 substitutions led to activation, rather than inhibition, of AKI by threonine. The otherreplacements conferred threonine insensitivity on AK!. The threonine sensitivity of HDI was also changed bythe amino acid substitutions at Ser-352. The HDI carried by the Tyr-352 mutant AKI-HDI was activated bythreonine. Single amino acid replacements at Ser-352 by Ala, Asn, Gln, His, Phe, Pro, Thr, or Tyr wereintroduced into truncated AKI-HDIs containing the AKI and the central regions. The AKI activity of thetruncated AKI-HDI containing the first 468 amino acid residues was sensitive to threonine, and introductionof the amino acid replacements did not alter the threonine sensitivity of the AK1. Another truncated AKI-HDIcontaining the first 462 amino acid residues possessed threonine-resistant AKI, whereas the substitutions ofSer-352 with Ala and Pro rendered AKI sensitive to threonine. The replacement of Gln-351 with Phe activatedAKI of the truncated AKI-HDI in the presence of L-threonine. These findings suggest that Ser-352 of the centralregion of AKI-HDI is possibly a key residue involved with the allosteric regulation of both AKI and HDIactivities.

Threonine is synthesized from L-aspartate by five enzy-matic reactions. In Serratia marcescens, four of five enzymeactivities of this pathway are encoded by the thrAlAfBCoperon (17, 24) as in Escherichia coli (4, 8, 19). The first andthird steps of this pathway are catalyzed by aspartokinase I(AKI; EC 2.7.2.4) and homoserine dehydrogenase I (HDI;EC 1.1.1.3), respectively, both of which are encoded by thethrAA2 genes in a single polypeptide chain, AKI-HDI. BothAKI and HDI activities are allosterically inhibited by thre-onine, and expression of the thr operon is subject to multi-valent repression by threonine and isoleucine. The AKI-HDIprotein, therefore, is a key enzyme of threonine synthesis inS. marcescens. We have previously reported the nucleotidesequence of the thr operon of S. marcescens and multimerformation of the AKI-HDI protein (17). The S. marcescensAKI-HDI protein is a large tetrameric polypeptide carryingthe AKI activity in the N-terminal region and the HDIactivity in the central and C-terminal regions. The centralregion is also responsible for multimer formation of thisprotein.

Threonine-producing strains of S. marcescens Sr4l, lack-ing feedback inhibition of HDI or of both AKI and HDI,have been obtained. Strains HNr21 and HNr59, carrying thethrA1lA21 and thrA1+A22 mutations, respectively, are resis-tant to a threonine analog, P-hydroxynorvaline (12). StrainTLr156 was isolated as a threonine-resistant strain frommutants which were sensitive to threonine owing to lack of

* Corresponding author.

homoserine dehydrogenase II and contains the thrA1SA25mutation. These strains produce 4 to 24 g of L-threonine perliter in a fermentation medium containing sucrose and urea(10-12, 24). Both AKI and HDI of the strains HNr21 andTLr156 are desensitized for threonine-mediated feedbackinhibition (10, 24). Strain HNr59 carries threonine-sensitiveAKI and threonine-resistant HDI (12).The mutations of these threonine-producing strains have

been revealed to result from a single amino acid substitutionin the AKI-HDI protein (17). Threonine-resistant AKI-HDIsencoded by the thrAl1A21 and thrA15A25 genes containedsingle amino acid substitutions of Gly-330 with Asp and ofSer-352 with Phe, respectively, in the central region ofAKI-HDI. The thrAj+A22 mutation, rendering HDI threo-nine-resistant, was a single amino acid substitution of Ala-479 with Thr in the HDI region of AKI-HDI. The centralregion of AKI-HDI, therefore, is closely related to theallosteric response of both AKI and HDI.Among these three substitutions, the amino acid replace-

ment of Ser-352 with Phe is the most effective for thedesensitization for threonine-mediated feedback inhibition.This paper deals with the changes in the allosteric responseof AKI and HDI activities resulting from amino acid substi-tutions at position 352 in the AKI-HDI protein.

MATERUILS AND METHODS

Bacterial strains and plasmids. The strains and plasmidsused are listed in Table 1. E. coli JM109 was used as a host

959

on March 24, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

960 OMORI AND KOMATSUBARA

TABLE 1. E. coli strains and plasmids

Strain, phage, Relevant characteristic(s)a Source oror plasmid enreference

StrainsJM109 A(lac-proAB) thi recA1/F'(lacPq lacZAM15) 29490A thrA met leu recA rK mK 23GT16 thrA1 metLM IysC pro ser 25MM294 thi-1 1BW313 HfrKL16PO/45[1ys(61-62)/dut-1 nug-1 thi-J rel-1] 14BMH71-18mutS A(lac-proAB) thi mutS215::TnlO(Tetr)/F'(laclq lacZAM15) 13MV1184 A(lac-proAB) thi A(srI-recA)306::Tn1O(Tetr)/F'(laclq lacZAM15) 26

Bacteriophage M13KO7 26PlasmidspWT201 pBR322::thrA1+A2+BC 17pUC119 Apr lacZa 26pWT29-119 pUC119::thr'Aj+A2+B' This workpUC19 Apr lacZa 29pWT211 pLG339::thrBCA(thrAj A2 ) 17pTHRWA6 pUC1981::thrAj+A2' 17pTHRM156 pUU1981::thrA15A25 17pTHRAD104 pUC1981::thrA$A2' 17pTHRAD105 pUC1981::thrA42' 17a Symbols used for relevant genotypes and phenotypes are as follows: thrA15, lack of feedback inhibition of aspartokinase I, and thrA25, lack of feedback

inhibition of homoserine dehydrogenase I. For additional details concerning plasmid construction, consult Materials and Methods.

for the construction of plasmids. The rich media LB, 2x TY,and SOC were prepared as described previously (21). Theminimal medium of Davis and Mingioli (5) was modified byomitting sodium citrate and supplemented with 0.5% glucoseas a carbon source. Required L-amino acids were used at aconcentration of 1 mM. Ampicillin was added at a concen-tration of 200 ,ug/ml.

Genetic methods and DNA sequencing analysis. DNA ma-nipulations were carried out according to standard proce-dures (21). The DNA sequence of the plasmids containingthe mutant thrA gene was determined by the dideoxy chaintermination method (22). Ordered deletion subclones wereobtained by unidirectional digestion (29). The DNA chainswere labeled with [a-32P]dCTP (400 Ci/mmol; 15 TBq/mmol).

Site-directed mutagenesis of the thrAlA2 genes. Oligonucle-otide-directed site-specific mutagenesis was performed forthe amino acid substitutions according to the method ofKunkel et al. (14). A 2.5-kb EcoRI fragment of pWT201,carrying the thrA gene, was recloned into the EcoRI-di-gested pUC119 to give pWT29-119, which was constructedto produce a single-stranded DNA. E. coli BW313 harboringpWT29-119 was grown at 37°C in 3 ml of 2x TY and infectedby the helper phage M13K07. After overnight culture,single-stranded uracil-containing phagemid DNA was iso-lated.

Oligonucleotides, 5'-'TTCGGAAGAxxxTITGGGTGATC-3'(xxx corresponds to the antisense codon of the amino acidreplaced at position 352), were synthesized for the Ser-352(TCC)-to-Ala (GCC), -Arg (AGG), -Asn (AAC), -Gln (CAG),-Glu (GAA), -His (CAC), -Leu (CTG), -Met (ATG), -Phe(TTC), -Pro (CCG), -Thr (ACG), -Trp (TGG), -Tyr (TAC),and -Val (GTC) mutations. A synthetic DNA, 5'-TATTCGGAAGAGGAGAAGGTGATCAGCA-3', was used for themutation from Gln-351 to Phe. The oligonucleotides werephosphorylated by T4 polynucleotide kinase prior to theiruse as primers in the synthesis reaction. Mutagenesis reac-tions were performed by using components available in theMutan-K site-directed mutagenesis kit (Takara Shuzo, Ky-

oto, Japan) and the procedure described in the instructionmanual for the kit. The reaction products (3 RI) weretransformed into competent cells (30 ,ul) of E. coli ung+BMH71-18mutS, which lacks a mismatch repair system.After 1 h of shaking at 37°C with 0.3 ml of SOC medium, thecells of BMH71-18mutS were infected by the helper phageM13KO7. The mutant plasmid DNA was recovered from theE. coli MV1184 cells which were infected by the resultantM13 phage. The mutations obtained were confirmed byDNA sequencing.Assay of enzyme activity. Plasmids containing the mutant

thrA gene were introduced into E. coli GT16 (thrA1 metLMlysC) for the AKI assay and into MM294 (thi) for the HDIassay. Cells were grown at 25°C for GT16 and at 30°C forMM294 in test tubes containing 3 ml ofLB medium including0.01% diaminopimelic acid and ampicillin. The cell extractswere prepared and used for the assays of AKI and HDIactivities as described previously (9). AKI activity wasmeasured by the formation of aspartylhydroxamate, andHDI activity was measured by the reduction of NADP.Protein concentration was determined by the use of aBio-Rad protein assay kit with bovine serum albumin as thestandard protein.Western blot (immunoblot) analysis of the gene products.

Cell lysates prepared from E. coli GT16 carrying the thrAplasmids were subjected to sodium dodecyl sulfate (SDS)-12.5% polyacrylamide gel electrophoresis (PAGE) usingthe buffer system of Laemmli (16). The proteins wereelectrophoretically transferred to an Immobilon P filter(Millipore). Immunostaining of the blots was performed by aperoxidase procedure (Vectastain ABC kit; Vector Labora-tories, Burlingame, California) using a 1:500-to-1:1,000 dilu-tion of the antiserum to a synthetic peptide, AKP (NH2-Asp-Arg-Leu-Lys-Gly-Val-Val-Asp-Gln-Glu-Phe-Ala-Gln-Leu-Lys-COOH), corresponding to amino acid residues 88 to 102of AKI-HDI (17).

Chemicals. Restriction enzymes, modification enzymes,and the DNA sequencing system were products of TakaraShuzo. [ot-32P]dCTP for DNA sequencing was supplied by

J. BACTERIOL.


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

CHANGES IN ALLOSTERIC RESPONSE OF AKI-HDI BY MUTATION 961

TABLE 2. Plasmids obtained by site-directed mutagenesis

Amino acid EnzymePlasmid residues Amino acid activity'

encoded substitutionAKI HDI

pTHRWA6 1-819 Ser-352b 170 190pTHRM156 1-819 Phe-352c 310 81pTHRM352A 1-819 Ser-352- Ala 89 510pTHRM352E 1-819 Ser-352--Glu 72 160pTHRM352H 1-819 Ser-352--His 7 3pTHRM352L 1-819 Ser-352--+Leu 400 130pTHRM352M 1-819 Ser-352---Met 120 350pTHRM352N 1-819 Ser-352--)Asn 200 310pTHRM352P 1-819 Ser-352-->Pro 160 210pTHRM352Q 1-819 Ser-352--Gln 180 210pTHRM352R 1-819 Ser-352--*Arg 140 470pTHRM352T 1-819 Ser-352--*Thr 160 290pTHRM352V 1-819 Ser-352--)-Val 190 190pTHRM352W 1-819 Ser-352--Trp 140 220pTHRM352Y 1-819 Ser-352--+Tyr 94 190pTHRM351F 1-819 Gln-351- Phe 7 3

pTHRAD104 1-468 Ser-352b 220pTD104-352A 1-468 Ser-352-->Ala 250pTD104-352F 1-468 Ser-352--*Phe 240pTD104-352H 1-468 Ser-352-*His 230pTD104-352N 1-468 Ser-352--Asn 210pTD104-352P 1-468 Ser-352- Pro 100pTD104-352Q 1-468 Ser-352-*Gln 180pTD104-352T 1-468 Ser-352--.Thr 180pTD104-352Y 1-468 Ser-352--*Tyr 210pTD104-351F 1-468 Gln-351--*Phe 230

pTHRAD105 1-462 Ser-352b 250pTD105-352A 1-462 Ser-352- Ala 380pTD105-352F 1-462 Ser-352--Phe 320pTD105-352H 1-462 Ser-352--His 120pTD105-352N 1-462 Ser-352- Asn 250pTD105-352P 1-462 Ser-352--Pro 270pTD105-352Q 1-462 Ser-352-*Gln 370pTD105-352T 1-462 Ser-352- Thr 320pTD105-352Y 1-462 Ser-352- Tyr 510pTD105-351F 1-462 Gln-351-->Phe 140

' Activities of AKI and HDI are expressed as micromoles of product perminute per milligram of protein and were assayed in the absence of L-threo-nine.

b Wild type.c thrA1SA25.

Amersham International (Little Chalfont, Buckinghamshire,United Kingdom). Other reagents were purchased fromKatayama Chemical (Osaka, Japan).

RESULTS

Site-directed mutagenesis of the thrA gene and constructionof thrA plasmids. The 0.8-kb BglII-digested fragment wasprepared from mutant plasmids carrying single amino acidsubstitutions at amino acid residue 352 and was inserted intothe BglII site of pWT'211, lacking the corresponding frag-ment of the thrA gene. Plasmids carrying the mutant thrAgene were selected as Thr+ clones of E. coli 490A (thrA) andwere confirmed to possess the functional thrA gene. The0.9-kb EcoRI-SphI-digested fragment, prepared from theseplasmids, was subcloned into the corresponding sites of thethrA plasmids, pTHRWA6, pTHRAD104, and pTHRAD105,for high-level expression of the AKI-HDI proteins. Themutations of the plasmids obtained were confirmed by DNAsequencing and are summarized in Table 2.Feedback inhibition of the mutant AKI-HDI enzyme con-

taining all 819 amino acid residues. Figure 1 shows theallosteric response of 12 mutant AKIs and HDIs containingsingle amino acid substitutions. We failed to detect both AKIand HDI activities of His-352 and Phe-351 mutant AKI-HDIenzymes. The AKIs of the Pro-352 and Thr-352 mutantenzymes were sensitive to threonine, as was that of the wildtype (Ser-352). Eight mutations, that is, single amino acidreplacements of Ser-352 of the wild type with Ala, Arg, Gln,Glu, Leu, Met, Trp, and Val, released AKI from threonine-mediated allosteric inhibition, as did the thrA1SA25 mutation(Phe-352). Interestingly, two substitutions including Asn-352and Tyr-352 elevated AKI activity twofold by the addition of50 mM L-threonine.The threonine sensitivity of HDIs was reduced by the

introduction of single amino acid substitutions at amino acidresidue 352. The Tyr-352 replacement, which activated AKIactivity in the presence of L-threonine, also increased HDIactivity about twofold by the addition of 100 mM L-threo-nine. While the Asn-352 replacement conferred threonine-mediated activation on AKI, this substitution desensitizedHDI for threonine-mediated feedback inhibition and did not

-- pTHRWA6 (Ser-352),, pTHRM352A (Ala-352)

200 - -'?A pTHRM352R (Arg-352)-A - pTHRM352N (Asn-352)

- .., -El- pTHRM352Q (Gln-352)-#--IpUpTHRM352E (Glu-352)Z

---O- pTHRM352L (Leu-352)

.2 p>ss > | \Q , \; . ' ---*-- THRM352M (Met-352)--I.\ ------ pTHRM156 (Phe-352)

,i A.B ---*-- pTHRM352P (Pro-352)A----- pTHRM352T (Thr-352)0u- . . , . . , . . ... , | |--R-- pTHRM352W (Trp-352)0 1 10 100 0 .1 1 10 100 _ pTHRM352Y (Tyr-352)

L-threonine added (mM) L-threonine added (mM) --+-- pTHRM352V (Val-352)FIG. 1. Threonine-mediated feedback inhibition of the AKI-HDI enzymes carrying amino acid replacements at Ser-352. Cell extracts were

prepared from cells carrying the wild type and the mutant thrA plasmids which were derived from pTHRWA6 coding for amino acid residues1 to 819. For the assays of AKI and HDI activities, E. coli GT16 (thrAl metLM lysC) and E. coli MM294 were used as hosts, respectively.Panels A and B show relative activities of AKI and HDI activities, respectively. The amino acid substitutions at residue 352 are shown atright. The activities in the absence of L-threonine are represented by 100%. Absolute values of these activities are listed in Table 2.

VOL. 175, 1993


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


(kDa)94-67-

43-r

301-

20

1 2 3 4 5

-AKI-HDI

-44kDa

FIG. 2. Analysis of the gene products by Western blotting. Thecell lysates of E. coli GT16 containing plasmids were subjected toSDS-12.5% PAGE and transferred to a filter. The blot was stainedby a peroxidase procedure. Lane 1, pUC19; lane 2, pTHRWA6; lane3, pTHRM156; lane 4, pTHRM352H; lane 5, pTHRM351F. Thepositions of the AKI-HDI protein and the 44-kDa protein are shownat right.

activate HDI by the addition of L-threonine. Since we didnot find the differences in thermostability between the mu-tant AKI-HDI enzymes (data not shown), these changes inthe allosteric response were not due to the stabilization ofAKI-HDI by the addition of L-threonine. Thus, these singleamino acid substitutions for Ser-352 altered the allostericresponse for threonine of AKI-HDI.

Analysis of expression of the His-352 and Phe-351 mutantAKI-HDIs in E. coli. Two replacements, His-352 and Phe-351, led to a loss of both AKI and HDI activities. To

120

100

. 80

0 60a).> 40

D 20

examine whether these enzymes were inactive or whethertheir expression was low in the cells, the gene productsencoded by pTHRM352H and pTHRM351F were analyzedby Western blotting using antiserum to AKI-HDI (Fig. 2).The 88-kDa thrA gene product, corresponding to the entireAKI-HDI protein, was observed in the cell lysate of E. coliGT16 carrying pTHRWA6 (wild type) and pTHRM156 (Phe-352 mutant). The 88-kDa gene product was not detected inthe cells carrying pTHRM352H and pTHRM351F, however.A weak immunoreactive 44-kDa band, which was not de-tected in the cells carrying pTHRWA6 and pUC19, wasobserved in the cells carrying pTHRM351F. Because theHis-352 and Phe-351 mutant AKI-HDIs had no activities, aproteolytic degradation of the AKI-HDI enzymes may haveoccurred.

Allosteric response for threonine of AKI activities of thetruncated mutant AKI-HDIs. Single amino acid substitutionsfor Ser-352 or Gln-351 were introduced into the truncatedAKI-HDI proteins lacking the HDI domain. PlasmidpTHRAD104, coding for the first 468 amino acid residues ofthe AKI-HDI protein, encodes threonine-sensitive AKI (17).Nine mutant plasmids were constructed, and the cells car-rying these plasmids were examined for AKI activities in thepresence of various concentrations of L-threonine (Fig. 3).All of the AKI activities carried by these truncated mutantAKI-HDIs were as sensitive to threonine as that encoded byplasmid pTHRAD104 (Ser-352), and their allosteric re-sponses for threonine were different from that of the wild-type AKI-HDI of 819 amino acid residues.Another truncated AKI-HDI, containing the first 462

amino acid residues, encoded by pTHRAD105, carries thre-onine-resistant AKI (17). Nine mutations were introducedinto the thrA gene of pTHRAD105, and then threonine-mediated feedback inhibition of AKIs encoded by the result-ant plasmids was examined (Fig. 4). Two replacements ofSer-352 with Ala and Pro rendered threonine-resistant AKIsensitive to threonine. AKI of the Phe-351 mutant wasactivated by the addition of 50 mM L-threonine. Threoninesensitivity of AKIs of other mutants was very similar to thatof AKI encoded by pTHRAD105. The His-352 and Phe-351replacements, leading to a loss of both AKI and HDIactivities of the AKI-HDI protein of 819 amino acid resi-

--0--0A

--U-

----0---

pTHRWA6 (wild-type)pTHRAD1 04 (Ser-352)pTD1 04-352A (Ala-352)pTD104-352N (Asn-352)pTD1 04-352Q (Gln-352)pTD1 04-352H (His-352)pTD1 04-352F (Phe-352)pTD1 04-352P (Pro-352)pTD1 04-352T (Thr-352)pTD1 04-352Y (Tyr-352)

----a--- pTD104-351 F (Phe-351)0 1 10 100

L-threonine added (mM)FIG. 3. Threonine-mediated feedback inhibition of AKI activity of the mutant AKI-HDI enzymes consisting of the first 468 amino acid

residues. Cell extracts were prepared from cells carrying the wild type and the mutant thrA1 plasmids which were derived from pTHRAD104.E. coli GT16 (thrA, metLM lysC) was used as a host. The amino acid substitutions are shown at right. The activities in the absence ofL-threonine are represented by 100%. Absolute values of these activities are listed in Table 2.

J. BACTERIOL.


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


0

0CZ1a)

Ca

--.-0--A-

0 .1 1 10L-threonine added (mM)

100

pTHRWA6 (wild-type)pTHRAD105 (Ser-352)pTD1 05-352A (Ala-352)

-i-- pTD105-352N (Asn-352)-El- pTD105-352Q (Gln-352)

pTD105-352H (His-352)----a--- pTD105-352F (Phe-352)----.--- pTD105-352P (Pro-352)----t--- pTD 105-352T (Thr-352)----A--- pTD105-352Y (Tyr-352)----o--- pTD105-351F (Phe-351)

FIG. 4. Threonine-mediated feedback inhibition of AKI activities of the mutant AKI-HDI enzymes consisting of the first 462 amino acidresidues. Cell extracts were prepared from cells carrying the wild type and the mutant thrAM plasmids which were derived from pTHRAD105.E. coli GT16 (thrA, metLM lysC) was used as a host. The amino acid substitutions are shown at right. The activities in the absence ofL-threonine are represented by 100%. Absolute values of these activities are listed in Table 2. The AKI activity encoded by pTHRAD105 isresistant to L-threonine, and the symbol for pTHRAD105 (Ser-352) (-) overlaps with other symbols.

dues, conferred AKI activities on the two truncated AKI-HDI enzymes containing the first 462 and 468 amino acidresidues.

DISCUSSION

Feedback inhibition of the AKI-HDI enzyme is one of theregulatory points of threonine biosynthesis. Threonine-pro-ducing mutants of S. marcescens possess threonine-resistantAKI-HDI. The thrAllA21 and thrA1SA25 mutations, carriedby the strains HNr21 and TLr156, have been revealed to besingle amino acid substitutions of Gly-330 with Asp and ofSer-352 with Phe, respectively (17). These mutations led tothe desensitization of both AKI-HDIs for feedback inhibi-tion by L-threonine and were located in the hydrophobiccentral region of the AKI-HDI protein (Fig. 5).

Single amino acid substitutions at position 352 of theAKI-HDI protein caused the drastic changes in the allostericproperties of both AKI and HDI activities. Amino acidresidue 352, therefore, is important for the allosteric regula-tion for threonine of the AKI-HDI protein. If this residue isinvolved in binding to threonine, single amino acid substitu-tions at position 352 presumably alter the allosteric proper-ties of AKI and HDI in the same way. Therefore, thehydrophobic region containing this residue is not involved inbinding of threonine to the allosteric site but rather isresponsible for the allosteric conformation leading to regu-lation of both AKI and HDI activities.The thrA1+A22 mutation, which is a single amino acid

substitution of Ala-479 with Thr, is located in the hydropho-bic HDI region containing the G-X-G-X-X-G motif respon-sible for HDI activity (27, 28). This mutation, leading to thedesensitization of HDI for threonine-mediated feedbackinhibition, does not alter the threonine sensitivity of AKI.The Pro-352 and Thr-352 mutations, which desensitized HDIfor feedback inhibition without changing the threonine sen-sitivity of AKI, are located in the hydrophobic central regionof AKI-HDI. Since these amino acid substitutions in the twopositions led to the same allosteric properties of the AKI-HDI enzyme, the hydrophobic interaction between the cen-tral region and the HDI region is possibly one of theimportant factors for allosteric regulation of HDI. The

allosteric properties of AKI and HDI are not related to thehydrophobicity of the amino acid replaced. The hydrophobicinteraction between the central region and the HDI region ispossibly involved in transmission of allosteric control bythreonine binding, and amino acid residue 352 is importantfor the local conformation of the hydrophobic central region.The allosteric response of AKI activities of the truncated

AKI-HDI protein (residues 1 to 468), lacking the HDIdomain, was not changed by the introduction of single aminoacid substitutions. However, in the truncated AKI-HDIpossessing the first 462 amino acid residues, the introductionof single amino acid substitutions at amino acid residues 351and 352 altered the threonine sensitivity of AKI. Since theabove-mentioned mutations in the hydrophobic central re-gion are important for the AKI allosteric response of thetruncated AKI-HDI lacking the HDI region, a hydrophobicinteraction between the central region and the AKI region ispossibly one of the important factors for the allostericresponse of AKI, as is the interaction between the centralregion and the HDI region.Although the His-352 and Phe-351 mutant thrA genes

complemented the thrA mutation of E. coli 490A, we failedto detect both AKI and HDI activities of these mutantAKI-HDI enzymes of 819 amino acid residues. Western blotanalysis showed degradation of the AKI-HDI protein carry-ing these mutations in the cells of E. coli. These mutations,therefore, cause the conformational change in the centralregion which allowed proteolytic digestion in the cells. Thus,amino acid residues 351 and 352 in the central region areimportant for the conformation of the AKI-HDI protein.The findings described above suggest that hydrophobic

regions of the AKI-HDI protein are involved in the allostericresponse of AKI and HDI activities. We compared theamino acid sequences of the AKI region with aspartokinasesequences and analyzed these sequences for hydropathy ofthis domain. In the sequences, there were found threeconserved regions: the N-terminal region, the region ofamino acid residues 200 to 230, and the hydrophobic centralregion. The N-terminal region was rather hydrophilic, andthe region of amino acid residues 200 to 230 had a hydro-phobic sequence. Since this region is located near the D-P-Rsequence (amino acid residues 236 to 238) homologous to

VOL. 175, 1993


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


0

GXGXXG

DPR II1330352 /Thr iI

Gly Ser Ala

AKI HDI

2 <~~2 JO 40 -600 800Amino acid residue number

CSYMI4KMIATO' KI~rTYYlSEfIfJANM .A...JAAVSKIGVtARflAXPW

266266260292212213268

Sinarcescens AKI-HIDI 175 AAAAIPADH T A EDIRLIKSKSYQUMeE.coli AKI-lIDI 175 AASRIPADHF oAATAGNEKfMVV NW AVR CpRMCC g EC PDIRLIKSMSYQ M3

E.coli AKIII 169 QLLPRLNEGN T*AKATg YGEAHHSSRVPm mVSA4KRjDEjAFANAWJS.cerevisiae AK 201 PFVSAKERIM.P.TFGLVPT-LNGNOGYILC(AL1VAYN!EL.KDoTFMAM PEWRLfDS2TPEE1SWIjB.subtilis AKII 121 LADQLEKGKo AILQGMTEDCGTTORGlETEVAIWAWKVIKCXYMPi.01| YjKSWRKbEGISYDIMLMC.glittaanicium AK 122 REALDEGKIC,AGFQVNKETRaTTMRGllTlVAWACNmVCEfYX IIPNIQKpEK#SFE#LE.coli AKII-HDII 177 QLLVQHPGKR-.",T&"l'ISRNNAG ; KVL DI PLOFIG. 5. Hydrophobicity profile of the S. marcescens AKI-HDI protein and comparison of the amino acid sequences of aspartokinase

enzymes. Analysis of hydrophobicity was performed by the method of Kyte and Doolittle (15); a span length of 12 amino acid residues wasused. Hydrophobic regions are indicated by a profile above the x axis. The open arrow indicates the domains of the AKI-HDI protein. Thecentral region (amino acid residues 300 to 470) is indicated by the crosshatched box. Similarities between AKI regions of S. marcescensAKI-HDI (17), E. coli AKI-HDI (8), E. coli AKIII (2), Saccharomyces cerevisiae AK (20), Bacillus subtilis AKII (3), Corynebacteriumglutamicum AK (7), and E. coli AKII-HDII (30) are indicated. Identical residues and substitutive alternations (I, L, V, M, and F; D and E;N and Q; R and K; and S and T), which are conserved more than 70% among these sequences, are shaded. The positions and amino acidsubstitutions of the thrAllA21, thrA15A25, and thrAj+A22 mutations are indicated above the open arrow. The G-X-G-X-X-G (27, 28) andD-P-R (18) motifs are shown in the figure, and the D-P-R sequences are boxed.

y-glutamyl kinase and responsible for the aspartokinaseactivity (18), the hydrophobic interaction between this re-

gion and the central region is probably involved in theallosteric regulation of AKI. A triglobular model for thepolypeptide chain of the intact AKI-HDI protein has beenproposed (6). These hydrophobic interactions among theAKI, HDI, and central domains may be possible factors inthe allosteric regulation of the AKI-HDI protein.

ACKNOWLEDGMENTS

We are grateful to G. N. Cohen for providing E. coli GT16 and forhis encouragement. We also thank I. Chibata and T. Tosa forencouragement.

REFERENCES1. Backman, K., M. Ptashne, and W. Gilbert. 1976. Construction

of plasmids carrying the CI gene of bacteriophage X. Proc. Natl.Acad. Sci. USA 73:4174-4178.

2. Cassan, M., C. Parsot, G. N. Cohen, and J. C. Patte. 1986.Nucleotide sequence of lysC gene encoding the lysin-sensitiveaspartokinase III of Escherichia coli K12. J. Biol. Chem.261:1052-1057.

3. Chen, N. Y., F. M. Hu, and H. Paulus. 1987. Nucleotidesequence of the overlapping genes for the subunits of Bacillussubtilis aspartokinase II and their control regions. J. Biol.Chem. 262:8787-8798.

4. Cossart, P., M. Katinka, and M. Yaniv. 1981. Nucleotidesequence of the thrB gene of E. coli and its two adjacent regions;the thrAB and thrBC junctions. Nucleic Acids Res. 9:339-347.

5. Davis, B. D., and E. S. Mingioli. 1950. Mutants of Escherichiacoli requiring methionine or vitamin B12. J. Bacteriol. 60:17-28.

6. Fazel, A., K. Muller, G. Le Bras, J. R. Garel, M. Wron, andG. N. Cohen. 1983. A triglobular model for the polypeptidechain of aspartokinase I-homoserine dehydrogenase I of Esch-erichia coli. Biochemistry 22:158-165.

7. Kalinowski, J., J. Cremer, B. Bachmann, L. Eggeling, H. Sahm,and A. Puhler. 1991. Genetic and biochemical analysis of theaspartokinase from Corynebacterium glutamicum. Mol. Micro-biol. 5:1197-1204.

8. Katinka, M., P. Cossart, L. Sibilli, I. Saint-Girons, M. A.Chalvignac, G. Le Bras, G. N. Cohen, and M. Yaniv. 1980.Nucleotide sequence of the thrA gene of Escherichia coli. Proc.Natl. Acad. Sci. USA 77:5730-5733.

9. Kisumi, M., S. Komatsubara, and I. Chibata. 1977. Enhance-ment of isoleucine hydroxamate-mediated growth inhibition andimprovement of isoleucine-producing strains of Serratia

3.0

0

>1

4-0L-

>I

-3.0

J. BACTERIOL.


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/


marcescens. Appl. Environ. Microbiol. 34:647-653.10. Komatsubara, S. 1987. Construction of amino acid-hyperpro-

ducing strains of Serratia marcescens by transductional andrecombinant DNA techniques, p. 757-765. In 0. M. Neijssel,R. R. van der Meer, and K. C. A. M. Luyben (ed.), Proc. 4thEur. Congr. Biotechnol. 1987, vol. 4. Elsevier, Amsterdam.

11. Komatsubara, S., M. Kisumi, and I. Chibata. 1979. Transduc-tional construction of a threonine-hyperproducing strain ofSerratia marcescens. Appl. Environ. Microbiol. 38:1045-1051.

12. Komatsubara, S., M. Kisumi, K. Murata, and I. Chibata. 1978.Threonine production by regulatory mutants of Serratiamarcescens. Appl. Environ. Microbiol. 35:834-840.

13. Kramer, B., W. Kramer, and H.-J. Fritz. 1984. Different base/base mismatches are corrected with different efficiencies by themethyl-directed DNA mismatch-repair system of E. coli. Cell38:879-887.

14. Kunkel, T. A., J. D. Roberts, and R. A. Zakour. 1987. Rapid andefficient site-specific mutagenesis without phenotypic selection.Methods Enzymol. 154:367-382.

15. Kyte, J., and R. F. Doolittle. 1982. A simple method fordisplaying the hydropathic character of a protein. J. Mol. Biol.157:105-132.

16. Laemmli, U. K. 1970. Cleavage of structural proteins during theassembly of the head of bacteriophage T4. Nature (London)227:680-685.

17. Omori, K., Y. Imai, S.-I. Suzuki, and S. Komatsubara. 1993.Nucleotide sequence of the Serratia marcescens threonineoperon and analysis of the threonine operon mutations whichalter feedback inhibition of both aspartokinase I and homoserinedehydrogenase I. J. Bacteriol. 175:785-794.

18. Omori, K., S. Suzuki, Y. Imai, and K. Komatsubara. 1991.Analysis of the Serratia marcescens proBA operon and feed-back control of proline biosynthesis. J. Gen. Microbiol. 137:509-517.

19. Parsot, C., P. Cossart, I. Saint-Girons, and G. N. Cohen. 1983.Nucleotide sequence of thrC and of the transcription termina-

tion region of the threonine operon in Escherichia coli K12.Nucleic Acids Res. 11:7331-7345.

20. Rafalski, J. A., and S. C. Falco. 1988. Structure of the yeastHOM3 gene which encodes aspartokinase. J. Biol. Chem.263:2146-2151.

21. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecularcloning: a laboratory manual, 2nd ed. Cold Spring HarborLaboratory, Cold Spring Harbor, N.Y.

22. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequenc-ing with chain-terminating inhibitors. Proc. Natl. Acad. Sci.USA 74:5463-5467.

23. Steinmetz, M., A. Winoto, K. Minard, and L. Hood. 1982.Clusters of genes encoding mouse transplantation antigens. Cell28:489-498.

24. Sugita, T., S. Komatsubara, and M. Kisumi. 1987. Cloning andcharacterization of the mutated threonine operon (thrA15A25BC) of Serratia marcescens. Gene 57:151-158.

25. Theze, J., D. Margarita, G. N. Cohen, F. Borne, and J. C. Patte.1974. Mapping of the structural genes of the three aspartoki-nases and of the two homoserine dehydrogenases of Escherichiacoli K-12. J. Bacteriol. 117:133-143.

26. Vieira, J., and J. Messing. 1987. Production of single-strandedplasmid DNA. Methods Enzymol. 153:3-11.

27. Wierenga, R. K., M. C. H. De Maeyer, and W. G. J. Hol. 1985.Interaction of pyrophosphate moieties with a-helixes in dinu-cleotide binding proteins. Biochemistry 24:1346-1357.

28. Wierenga, R. K., P. Terpstra, and W. G. J. Hol. 1986. Predictionof the occurrence of the ADP-binding Pap3-fold in proteins,using an amino acid sequence fingerprint. J. Mol. Biol. 187:101-107.

29. Yanisch-Perron, C., J. Vieira, and J. Messing. 1985. ImprovedM13 phage cloning vectors and host strains: nucleotide se-quences of the M13mpl8 and pUC19 vectors. Gene 33:103-119.

30. Zakin, M. M., N. Duchange, P. Ferrara, and G. N. Cohen. 1983.Nucleotide sequence of the metL gene of Escherichia coli. J.Biol. Chem. 258:3028-3031.

VOL. 175, 1993


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

marcescens aspartokinase 1-homoserine dehydrogenase i

Documents